Exhaust into Fuel: Turning Data into a Strategic Business Asset with AI

8 minute read

When you hear “artificial intelligence (AI)” do you think about bad robots like HAL and the Terminator? Don’t. Think R2D2 and C3PO. Like those two very useful droids, artificial intelligence will bring tremendous benefits to humans. AI and machine learning are here and in widespread use every day—in fact, we’ve already become quite dependent on them.

Some common examples of machine learning and AI include GPS or ride-sharing applications that approximate time of travel or the quickest route based on live traffic data; online shopping, which offers you a list of recommendations based on your previous purchases and preferences; and a learning thermostat, like Nest, which uses behavioral algorithms to customize your home’s temperature and energy usage.

It’s no coincidence that the examples cited above are also examples of digital businesses that have blossomed and then quickly dominated their industries. What did these businesses figure out before the rest? They learned how to use data science as a strategic business asset that fuels and promotes not only more business, but also smarter business.

In our digital world, the key to gaining a competitive advantage is learning to see data as currency. We are producing more data than ever—roughly 2.5 quintillion bytes of data every day—and that number is only expected to grow. While all this data may seem to be merely the exhaust or remnants of our digital trail, data scientists have been turning this data into much more, including the adoption of machine learning and artificial intelligence.

What Makes AI Real: IT Modernization and Cloud

So how is AI possible? IT modernization and cloud computing continue to play crucial roles in the development of data science and AI. The ability to leverage data stored in the cloud and data centers, along with the advancements in storage and networking, have made it possible for data scientists to advance the industry of data science. They can label, classify and tag data better, gaining the ability to ingest the data faster and analyze it for insight. This is known as the "virtuous data cycle,” or “virtuous cycle of growth,” in which edge capabilities generate data through improved high-speed networks going into the data center and the cloud. 

A commitment to AI adoption is also a commitment to the virtuous data cycle. It’s a strategy for growth for every enterprise, regardless of maturity. It starts with hyperconverged infrastructure, advanced storage and computational power in the data center and cloud, and continues to mature with the development of application-specific integrated circuits for Al workloads. 5G networks, which are right around the corner, will connect the data center and the cloud to a host of edge devices and the Internet of Things (IoT), supplying unprecedented amounts of data and computational power. Along with our ecosystem of partners, we look to accelerate training, inference and the adoption of AI.

In short, cloud computing makes AI and machine learning real. It drives an economic business model around computational and data storage capabilities. The expanding use of the hybrid cloud—private, public, community—to deploy IT services is ultimately driving AI capabilities, too.

Key Terms of Data Science

Core Analytics: Descriptive, diagnostic, predictive and prescriptive analytics that describe what happened, why it happened, what will happen next, and what should be done about it.

Machine Learning and Cognitive Computing: Use of computational algorithms to learn from the data we feed it without programming.

  • The goal of a training algorithm is to be as accurate as possible through the ingestion of loads of data.
  • The goal of an inference algorithm is to present new data not previously seen so the machine can infer what it is. (Examples: Jeopardy, Go, and Poker, where a machine learns the rules, ingests lots of games/data, and we let it play itself until eventually it can beat a human at the game.

Deep Learning: set of convolutional neural networks that allow for analytics in multiple granular stages with a higher and higher degree of accuracy of insight. Examples include image recognition and facial recognition.

Artificial Intelligence: the outsourcing of human cognition to the thinking machine that operates at internet speed. Examples include autonomous driving.

Enterprises Go Data-Centric

Every organization needs to consider the move to a data-centric enterprise. Why? Becoming data-centric means harnessing the power of data to become more competitive. Data can be used to generate profits or productivity, gain market insight, encourage innovation, or personalize services and healthcare. Data can be used to provide context to make wiser decisions, identify new opportunities for the business, and contribute knowledgeable analysis to every part of the business cycle.

The amount of data produced continues to grow exponentially from gigabytes to zettabytes with every search on Google, every upload to YouTube, and every post to Twitter. Enterprises feel the need and the pressure to innovate faster and smarter in the dizzying pace of the digital world—and that requires them to keep up with the advancement of data science and AI.

According to the study State of Artificial Intelligence for Enterprises, businesses are investing heavily in the future of AI: 80% already have some form of AI in production today; 30% plan to expand their AI investments over the next two to three years; and 62% plan to hire a Chief AI Officer in the future. However, heavy enterprise investment in AI comes with enormous expectations. Globally, companies investing in AI expect a $1.23 ROI in the next three years for every dollar invested, $1.99 ROI in the next five years, and $2.87 ROI in the next ten years.

Clearly the investments in AI and the high ROI forecast are expected to bring unprecedented change to businesses and perhaps even revolutionize our everyday lives. To imagine this AI-prominent future, we must turn to the best business advantage use cases.

Business Use Cases for AI

Business use cases for machine learning and AI span every industry. In the energy sector, AI could meet the global demand for low-carbon, green electricity by solving the problem of renewable energy’s inconsistency. Take wind power as an example. While a set of windless days may generate a power shortfall and a string of sunny, windy days could generate more energy than consumed, AI could be used to store back-up power in order to deliver renewable energy more consistently, regardless of the particular weather conditions of one day or one place.

Life science is another fast-moving area for AI applications. Wearables like Fitbit are ubiquitous today. In addition, the use of cloud has improved access to medical records and diagnoses. However, with the future use of genomics and biometrics, AI could provide a better understanding of an individual patient’s needs, resulting in personalized immunotherapy or the development of pharmaceuticals to treat the specific ailments of an individual. Another example of how AI is advancing health science is image recognition. New AI technologies are enabling earlier and better detection of cancer cells.

AI’s Biggest Challenges: The Talent Gap and Ethics

Today there is a significant shortage of data scientists. By 2020, there will be more than 1.4 million computing-related job openings in the U.S. alone. But at current rates, with the number of STEM-related college graduates in the U.S. falling, only about 30% of those jobs will be filled by U.S. graduates. In order to advance data science and AI capabilities, we need to match the number and quality of AI professionals. One way to effectively increase engagement with skills and talent is for companies to embed machine learning and AI programs within universities.

Intel collaborates with approximately 400 universities worldwide to create career-ready graduates and has invested more than $1 billion in global higher education. With our embedded university programs, we pursue a number of goals, from collaborating with professors and students by providing advanced technology, to encouraging students to pursue technical studies, to working with the government and NGOs to fund higher education initiatives.

Above all, we believe in the need to engage young people in data science and expand different perspectives and opportunities in the industry. Young people are, in fact, providing so much of the data themselves. Technology, combined with development imperatives such as critical thinking, entrepreneurial skills, and a strong foundation in math and science, empowers students to expand their individual impact and tackle global problems across the planet.

Certainly, another challenge to the adoption of AI is the ethical, legal, and societal implications. Issues of security, privacy and ethics are beginning to be addressed with new laws such as GDPR in the EU, which guarantees the right of an individual—and their data—to be forgotten. The organization Data for Democracy is currently working on creating a code of ethics for data scientists by data scientists that can address and hopefully prevent the weaponization of data or bias in algorithms.

The Future of AI

Machine learning and AI face a bright future of possibilities. Certainly, the widespread adoption of AI will depend on the growth of the core infrastructure needed to sustain it. A recent Pew study reports that 11% of Americans (that would be about 35 million) don’t use the internet—there is much more data and services to be consumed.

Another development in AI will be the rise of citizen data scientists. More people will begin to understand how to use data to their own benefit, not just the billion-dollar global companies. For example, if a person wants to make a case for installing a stoplight in a neighborhood intersection, they can access and analyze traffic data from the intersection and compare the number of incidents with similar intersections. If the data shows a much higher rate of accidents or speeding tickets than other intersections, the citizen data scientist has a validated argument for change.

Both the expansion of core infrastructure and the rise of the citizen data scientist also points to another inspiring prediction: more people using more data everywhere also means the expansion of the availability of benefits of AI to a more diverse population. In the future, AI will operate on a notion of inclusivity—turning data into a strategic asset for all.


[SlideShare] Exhaust into Fuel: Turning Data into a Strategic Business Asset

View more presentations from Forsythe Technology

You Might Also Like
Join our Newsletter

Stay up to date with the latest and greatest from our monthly newsletter

About the Author
Popular Today